Assessment of Neck Grasp as a Screening Tool for Identifying Obstructive Sleep Apnea in Community-Dwelling Older Adults

Abstract

Background

Inability to encircle the neck by hands (neck grasp) has been proposed as an indicator of obstructive sleep apnea (OSA) that would be useful for recognition of candidates for polysomnography (PSG). We assessed the value of neck grasp for predicting OSA in community-dwelling older adults of Amerindian ancestry.

Methods

Neck grasp was evaluated in individuals aged ≥60 years undergoing PSG. The association between neck grasp and OSA was assessed by logistic regression models adjusted for relevant covariates. Mediation analysis was used to establish the proportion of the effect of the association between neck grasp and OSA, which is mediated by the neck circumference (a well-known OSA biomarker). Receiver operator characteristics curve analysis was used to estimate diagnostic accuracy of neck grasp for predicting OSA.

Results

Of 201 individuals undergoing PSG, 167 (83%) had the neck grasp test. The remaining 34 could not perform the test because of different factors. Neck grasp was positive in 127 (76%) cases, and 114 (68%) individuals had OSA (apnea-hypopnea index ≥5). Multivariate logistic regression models disclosed a significant association between neck grasp and OSA. The neck circumference was the single covariate remaining independently significant in these models. Neck grasp was not efficient at predicting OSA (sensitivity: 83.3%, specificity: 39.6%, positive predictive value: 0.75 and negative predictive value: 0.53). The area under the curve disclosed only a moderate predictive capability (61.5%) of neck grasp.

Conclusion

Results do not support the use of neck grasp as an independent predictor of OSA in the study population.

Keywords

neck grasp polysomnography obstructive sleep apnea neck circumference Atahualpa Amerindians

Introduction

Obstructive sleep apnea (OSA) increases the risk of stroke and all-cause mortality.¹ According to conservative estimates, about 1 billion of adults worldwide have OSA.² However, the burden of this condition remains unknown in remote communities, mostly because technology needed for OSA diagnosis—polysomnography (PSG) is not readily available. Moreover, mass screening with field instruments designed to identify candidates for PSG, such as the Epworth Sleepiness Scale or the Berlin questionnaire, have shown a disappointing diagnostic accuracy due to cross-cultural factors and illiteracy.^3-5 In addition, several studies have evaluated the diagnostic accuracy of anthropometric indices in this setting. With the exception of the neck circumference, which has been almost consistently associated with OSA, the debate on the value of other anthropometric indices is never-ending and probably depends on several factors, including race/ethnicity and the environment.^6-14

There is a need for a screening tool that may be able to identify persons at risk for OSA in rural populations. In this view, 2 small series have shown 100% specificity of neck grasp for identifying subjects with OSA.^15,16 However, these results have not been replicated in other populations/ethnic groups. Taking the opportunity of a well-established cohort of community-dwellers living in rural Ecuador, this study aimed to assess the independent association between neck grasp and OSA as well as the diagnostic accuracy of neck grasp for predicting OSA in older adults of Amerindian ancestry.

Methods and Materials

This study was conducted in Atahualpa, a rural village located in coastal Ecuador, and approved by the Institutional Review Board of our Institution. All individuals signed a comprehensive informed consent before enrollment. As detailed elsewhere, homogeneity of the population regarding race/ethnicity (Amerindian ancestry) and lifestyles, as well as the low migration rate, provide an optimal setting for the conduction of epidemiological studies by reducing the occurrence of hidden or unexpected confounders.¹⁷

Neck grasp was evaluated in community-dwelling older adults aged ≥60 years undergoing PSG for research purposes.^18,19 Individuals undergoing PSG had been selected by the use of the Random Integer Generator (https://www.random.org/integers/), as previously described.¹⁸ The neck grasp consisted on instructing the subject to place the thumbs together in the ventral aspect of the neck and to wrap fingers around the neck until they met in the posterior aspect. Inability to touch fingertips is a positive test and the opposite is a negative test (Figure 1). Individuals unable to perform the neck grasp were excluded.

Figure 1.

The neck grasp is performed by instructing the subject to place the thumbs together in the ventral aspect of the neck and to wrap fingers around the neck until they met in the posterior aspect (without severe choking). The inability to touch fingertips is a positive test (left panel) and the ability to encircle the neck with fingers is a negative test (right panel).

As detailed elsewhere, single-night diagnostic PSGs were performed at the sleep unit of our Center. Exams were done with the use of an Embletta® X100™ Comprehensive Portable PSG System (Embla Systems, Inc; Thornton, CO, USA).¹⁸ A certified sleep medicine neurologist reviewed raw data and interpreted all exams blinded to clinical data.²⁰ Attention was given to the apnea-hypopnea index (AHI), which calculated the number of apnea/hypopnea events divided by the number of sleep hours. OSA was defined if the AHI was ≥5.²

Demographics (age and sex), relevant cardiovascular risk factors (body mass index, blood pressure, fasting glucose and total cholesterol blood levels), and the neck circumference were selected as clinical covariates. Demographics and cardiovascular risk factors were assessed through interviews and procedures previously described in the Atahualpa Project.²¹ The neck circumference was measured using a stretch-resistant tape with individuals on the seated position, immediately above the cricoid cartilage and perpendicular to the long axis of the neck. These covariates were selected as they have been shown to be related to the main variables investigated in the study population.¹⁸

Data analyses were carried out by using STATA version 16 (College Station, TX, USA). In univariate analyses, continuous variables (expressed as means ± SD) were compared by linear models (ANOVA) and categorical variables (expressed as percentages with 95% confidence intervals when needed) by the x² or Fisher exact test as appropriate. Multivariate models were fitted to assess the association between a positive neck grasp and OSA (as the dependent variable), after adjusting for the above-mentioned covariates. Mediation analysis was used to establish the proportion of the effect between neck grasp and OSA, which is mediated by the neck circumference.²² Sensitivity analysis assessed the diagnostic accuracy of a positive neck grasp to predict OSA. Using receiver operator characteristics curve analysis, we calculated the area under the curve (AUC) for the predictive value of a positive neck grasp and the neck circumference for classifying subjects with OSA as positive or negative.

Results

A total of 167 out of 201 (83%) community-dwellers aged ≥60 years undergoing PSG were included in this study. The remaining 34 individuals could not perform the test because of inability to understand the instructions (mostly related to dementia), history of a stroke causing arm weakness, recent fractures of the fingers or arms, or partial or complete shoulder ankylosis. The mean age of participants was 71.7 ± 6.9 years (median age: 70 years), 107 (64%) were women, 39 (23%) had a body mass index ≥30 kg/m², 78 (47%) had blood pressure ≥140/90 mmHg, 47 (28%) had fasting glucose ≥126 mg/dL, and 27 (16%) had total cholesterol blood levels ≥240 mg/dL. The mean (±SD) value of the neck circumference was 35.2 ± 2.9 cm (median value: 35 cm).

A total of 127 (76%) individuals had a positive neck grasp, and 114 68%) had OSA. Of 127 subjects with a positive neck grasp, 95 (75%) had OSA. On the other hand, of the 40 individuals with a negative neck grasp, only 19 (48%) had OSA (OR: 3.28; 95% C.I.: 1.57-6.07; P = .002). Also in univariate analyses, covariates associated with a positive neck grasp were a body mass index ≥30 kg/m², blood pressure levels ≥140/90 mmHg, and increasing neck circumference. In contrast, the only significant difference in the covariates investigated across individuals with and without OSA was a higher neck circumference in the former (Table 1).

Table 1.

Characteristics of Atahualpa Residents Included in This Study Across Categories of Neck Grasp and Obstructive Sleep Apnea (Univariate Analyses).

Variable	Total series (n = 167)	Neck grasp			Obstructive sleep apnea
Variable	Total series (n = 167)	Negative neck grasp (n = 40)	Positive neck grasp (n = 127)	P value	Apnea-hypopnea index <5 per hour (n = 53)	Apnea-hypopnea index ≥5 per hour (n = 114)	P value
Age, years (mean ± SD)	71.7 ± 6.9	70.9 ± 5.7	71.2 ± 7.3	.812	70.5 ± 7.2	71.4 ± 6.8	.436
Women, n (%)	107 (64)	25 (63)	82 (65)	.813	39 (74)	68 (60)	.081
Body mass index ≥30 kg/m², n (%)	39 (23)	1 (2.5)	38 (30)	<.001*	8 (15)	31 (27)	.085
Blood pressure ≥140/90 mmHg, n (%)	78 (47)	10 (25)	68 (53)	<.001*	22 (42)	56 (49)	.359
Fasting glucose ≥126 mg/dL, n (%)	47 (28)	8 (20)	39 (31)	.189	17 (32)	30 (26)	.441
Total cholesterol levels ≥240 mg/dL, n (%)	27 (16)	5 (13)	22 (17)	.469	9 (17)	18 (16)	.846
Neck circumference, cm (mean ± SD)	35.2 ± 3	33.8 ± 2.6	35.7 ± 2.9	<.001*	33.8 ± 2.9	35.9 ± 2.9	<.001*

Statistically significant result.

A multivariate regression model, including the above-mentioned covariates but the neck circumference, showed a significant association between a positive neck grasp and OSA (OR: 3.03; 95% C.I.: 1.33-6.88; P = .008). When the neck circumference was added as a covariate to the above-mentioned model, a fully-adjusted model disclosed that the significance of the association between a positive neck grasp and OSA vanished (OR: 2.29; 95% C.I.: 0.97-5.39; P = .059). However, the fully-adjusted model contained several covariates that were not related to the outcome (P > .6) and contributed little to the model. Then, we fitted a most parsimonious model that included relevant covariates, and found an independent significance between a positive neck grasp and OSA (OR: 2.36; 95% C.I.:1.06-5.28; P = .036). The neck circumference remained as the single covariate achieving independent significance in both the fully-adjusted and the most parsimonious models (Table 2). Mediation analysis demonstrated that 35% (95% C.I.: 21.3%-90.2%) of the effect of the association between neck grasp and OSA was mediated by the neck circumference.

Table 2.

Uni- and Multivariate Logistic Regression Models, Adjusted for Different Covariates, Using a Positive Neck Grasp as the Independent Variable and Obstructive Sleep Apnea (OSA) as the Outcome (Dependent Variable).

Model	Covariates included	Significance of covariates	Significance between neck grasp and OSA
Univariate	None	. . .	OR: 3.28; 95% C.I.: 1.57 to 6.07; P = .002*
Multivariate logistic regression model (without neck circumference)	Age	OR: 1.04; 95% C.I.: 0.98 to 1.09; P = .192	OR: 3.03; 95% C.I.: 1.33 to 6.88; P = .008*
	Female gender	OR: 0.37; 95% C.I.: 0.17 to 0.82; P = .015*
	Body mass index ≥30 kg/m²	OR: 2.33; 95% C.I.: 0.88 to 6.14; P = .088
	Blood pressure ≥140/90 mmHg	OR: 0.98; 95% C.I.: 0.47 to 2.07; P = .963
	Fasting glucose ≥126 mg/dL	OR: 0.57; 95% C.I.: 0.26 to 1.25; P = .160
	Total cholesterol ≥240 mg/dL	OR: 1.04; 95% C.I.: 0.40 to 2.69; P = .934
Fully-adjusted logistic regression model	Age	OR: 1.04; 95% C.I.: 0.98 to 1.10; P = .133	OR: 2.29; 95% C.I.: 0.97 to 5.39; P = .059
	Female gender	OR: 1.01; 95% C.I.: 0.36 to 2.88; P = .981
	Body mass index ≥30 kg/m²	OR: 1.15; 95% C.I.: 0.38 to 3.45; P = .806
	Blood pressure ≥140/90 mmHg	OR: 0.99; 95% C.I.: 0.46 to 2.13; P = .975
	Fasting glucose ≥126 mg/dL	OR: 0.51; 95% C.I.: 0.23 to 1.15; P = .104
	Total cholesterol ≥240 mg/dL	OR: 1.08; 95% C.I.: 0.41 to 2.84; P = .883
	Neck circumference	OR: 1.32; 95% C.I.: 1.09 to 1.59; P = .003*
Most parsimonious logistic regression model	Age	OR: 1.04; 95% C.I.: 0.99 to 1.09; P = .126	OR: 2.36; 95% C.I.: 1.06 to 5.28; P = .036*
	Fasting glucose ≥126 mg/dL	OR: 0.52; 95% C.I.: 0.23 to 1.14; P = .102
	Neck circumference	OR: 1.32; 95% C.I.: 1.14 to 1.53; P < .001*

Statistically significant result.

The proportion correctly classified for a positive neck grasp in predicting OSA was moderate (69.5%), while the sensitivity was high 83.3% [74.9-89.4%], and specificity low 39.6% [26.7-53.9%]. The AUC for a positive neck grasp in identifying subjects with OSA was 0.615 (0.539-0.689) and that of the neck circumference was 0.683 (0.597-0.769).

Discussion

This study shows a significant association between a positive neck grasp and OSA in community-dwelling older adults of Amerindian ancestry. However, mediation analysis disclosed that roughly one-third of this association is mediated by the neck circumference, probably explaining the poor diagnostic accuracy found in the sensitivity analysis and the mild predictive value (61.5%) of the AUC.

As previously noticed, only 2 studies have addressed the value of neck grasp for predicting OSA.^15,16 Both studies were conducted by the same group of investigators. The age of participants in 1 of them ranged from 29 to 81 years, and the other only mentioned that participants were aged 18 years or older. Race/ethnicity was not specified. Studies are limited by the small number of cases included (47 and 43 patients, respectively) and by the fact that 1 of them was confined to patients with diabetes mellitus. There was no information on the percentage of eligible candidates unable to perform the maneuver. Using an AHI cutoff ≥5 to diagnose OSA, both studies found a specificity of 100% of a positive neck grasp for predicting OSA, and a sensitivity ranging from 55% to 68.3%. According to the authors, neck grasp is easy to perform and will serve as an unequivocal identifier of subjects with OSA. However, it was recognized the sizable percentage of false negative cases, and the fact that they enrolled a selected population attending a sleep clinic, which may not necessarily be extrapolated to the population at large.

Subjects participating in the current study were identified by means of door-to-door surveys and randomly selected to undergo PSG for research purposes.^18,19 This reduces the possibility of selection bias. In such population, neck grasp did not perform well for predicting OSA despite the significant association found between the main variables investigated, although the AUC statistic may not reflect the discreteness of the data.²³

Likely, the effect of the neck circumference was responsible for this poor diagnostic accuracy, as shown in the above-described models. In addition, 17% of individuals could not preform the test because of several reason already described. This is a major limitation of the test if it is planned to be used for mass screening of OSA.

Major strengths of our study include the unbiased selection of participants taken from the community and the models used to assess diagnostic accuracy of neck grasp for predicting OSA. Limitations of the study are those of the neck grasp maneuver itself, that is, correct performance of neck grasp, which depends on proper understanding of the maneuver, integrity of shoulder girdle articulations, and other factors. It is also possible that some hidden confounders may have influenced the results of this study. In addition, the study population is limited to individuals of Amerindian ancestry, and our results may not be generalized to other races/ethnic groups.

Study results do not support the use of neck grasp as an independent predictor of OSA. Neck circumference is easier to measure and provide a more accurate information on the OSA status, as previously shown.¹⁴ More information is needed before widespread acceptance of neck grasp as a mass screening tool for predicting OSA in different populations/ethnic groups.

Footnotes

Authors’ Contributions

OHD: study design, manuscript drafting; RMM: statistical analyses; BYR: data collection and analysis; PRC: reading and interpretation of polysomnograms, significant intellectual contribution to manuscript content.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Study supported by Universidad Espíritu Santo—Ecuador. The sponsor had no role in the design of the study, nor in the collection, analysis, and interpretation of data.

ORCID iD

Oscar H. Del Brutto

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

Wang

Ouyang

Wang

Zhao

Liu

Obstructive sleep apnea and risk of cardiovascular disease and all-cause mortality: a meta-analysis of prospective cohort studies. Int J Cardiol. 2013;169:207-214.

Benjafield

Ayas

Eastwood

, et al. Estimation of the global prevalence and burden of obstructive sleep apnoea: a literature-based analysis. Lancet Respir Med. 2019;7:687-698.

Pahwa

Karunanayake

Hagel

, et al. Prevalence of High Epworth Sleepiness Scale scores in a rural population. Can Respir J. 2012;19:e10-e14.

Ulasli

Gunay

Koyuncu

, et al. Predictive value of Berlin Questionnaire and Epworth Sleepiness Scale for obstructive sleep apnea in a sleep clinic population. Clin Respir J. 2014;8:292-296.

Del Brutto

Zambrano

Castillo

PR.

Poor reliability of sleep questionnaires for the detection of sleep disordered breathing in a rural setting (The Atahualpa Project). Sleep Med. 2017;32:272.

Moraes

Povares

Zalcman

, et al. Association between body mass index and sleep duration assessed by objective methods in a representative sample of the adult population. Sleep Med. 2013;14:312-318.

Soylu

Levent

Sanman

Yurtlu

Alparslan

Saygi

Obstructive sleep apnea syndrome and anthropometric obesity indexes. Sleep Breath. 2012;16:1151-1158.

Kawaguchi

Fukumoto

Inaba

, et al. Different impacts of neck circumference and visceral obesity on the severity of obstructive sleep apnea syndrome. Obesity (Silver Spring). 2011;19:276-282.

Moul

Krishna

Neck circumference-height ratio as a predictor of sleep related breathing disorder in children and adults. J Clin Sleep Med. 2016;12:311-317.

10.

Cho

Choi

Suh

Ryu

Cho

SH.

Comparison of anthropometric data between Asian and Caucasian patients with obstructive sleep apnea: a meta-analysis. Clin Exp Otorhinolaryngol. 2016;9:1-7.

11.

Unal

Ozturk

Tosun

Kutlu

Association between obstructive sleep apnea syndrome and waist-to-height ratio. Sleep Breath. 2019;23:523-529.

12.

Ciavarella

Tepedino

Chimenti

, et al. Correlation between body mass index and obstructive sleep apnea severity indexes—a retrospective study. Am J Otolaryngol. 2018;39:388-391.

13.

Kang

, et al. The associations between anthropometric indices and obstructive sleep apnea in a Korean population. PLoS One. 2014;9:e114463.

14.

Del Brutto

Mera

Castillo

PR.

Anthropometric measurements in community-dwelling older adults of Amerindian ancestry with obstructive sleep apnea. The Atahualpa project. Sleep Med. 2020;68:18-19.

15.

Edmonds

LC.

A pilot study of the inability to fit hands around the neck as a predictor of obstructive sleep apnea. N Am J Med Sci. 2015;7:553-557.

16.

Edmonds

Gunasekaran

Edmonds

LC.

Neck grasp predicts obstructive sleep apnea in type 2 diabetes mellitus. Sleep Disord. 2019;2019:3184382.

17.

Del Brutto

Mera

Castillo

Del Brutto

VJ.

Key findings from the Atahualpa Project: what should we learn?

Expert Rev Neurother. 2018;18:5-8.

18.

Del Brutto

Mera

Zambrano

Castillo

PR.

Relationship between obstructive sleep apnea and neuroimaging signatures of cerebral small vessel disease in community-dwelling older adults. The Atahualpa Project. Sleep Med. 2017;37:10-12.

19.

Del Brutto

Mera

Costa

Castillo

PR.

Effect of heart rate variability on the association between the apnea-hypopnea index and cerebral small vessel disease. Stroke. 2019;50:2486-2491.

20.

Berry

Budhiraja

Gottlieb

, et al. Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events. Deliberations of the Sleep Apnea Definitions Task Force of the American Academy of Sleep Medicine. J Clin Sleep Med. 2012;8:597-619.

21.

Del Brutto

Santamaría

Ochoa

, et al. Population-based study of cardiovascular health in Atahualpa, a rural village of coastal Ecuador. Int J Cardiol. 2013;168:1618-1620.

22.

VanderWeele

TJ.

Mediation analysis: a practitioner’s guide. Annu Rev Public Health. 2016;37:17-32.

23.

Muschelli

III . ROC and AUC with a binary predictor: a potentially misleading metric. J Classif. 2020;37:696-708.